Plant Biotechnology Journal (2016) 14, pp. 195–205
doi: 10.1111/pbi.12373
Bottlenecks in carotenoid biosynthesis and accumulation
in rice endosperm are influenced by the precursor–
product balance
Chao Bai1, Teresa Capell1, Judit Berman1, Vicente Medina1, Gerhard Sandmann2, Paul Christou1,3 and
Changfu Zhu1,*
1
Department of Plant Production and Forestry Science, ETSEA, University of Lleida-Agrotecnio Center, Lleida, Spain
2
Institute of Molecular Bioscience, J. W. Goethe University, Frankfurt am Main, Germany
Institucio Catalana de Recerca i Estudis Avancats, Passeig Lluıs Companys, Barcelona, Spain
3
Received 6 December 2014;
revised 23 February 2015;
accepted 2 March 2015
*Correspondence (Tel +34 973702694;
fax 0034-973-702515; email zhu@pvcf.
udl.cat)
Keywords: rice (Oryza sativa),
carotenoids, secondary metabolites,
1-deoxy-D-xylulose 5-phosphate
synthase, ORANGE gene, multigene
engineering.
Summary
The profile of secondary metabolites in plants reflects the balance of biosynthesis, degradation
and storage, including the availability of precursors and products that affect the metabolic
equilibrium. We investigated the impact of the precursor–product balance on the carotenoid
pathway in the endosperm of intact rice plants because this tissue does not normally accumulate
carotenoids, allowing us to control each component of the pathway. We generated transgenic
plants expressing the maize phytoene synthase gene (ZmPSY1) and the bacterial phytoene
desaturase gene (PaCRTI), which are sufficient to produce b-carotene in the presence of
endogenous lycopene b-cyclase. We combined this mini-pathway with the Arabidopsis thaliana
genes AtDXS (encoding 1-deoxy-D-xylulose 5-phosphate synthase, which supplies metabolic
precursors) or AtOR (the ORANGE gene, which promotes the formation of a metabolic sink).
Analysis of the resulting transgenic plants suggested that the supply of isoprenoid precursors
from the MEP pathway is one of the key factors limiting carotenoid accumulation in the
endosperm and that the overexpression of AtOR increased the accumulation of carotenoids in
part by up-regulating a series of endogenous carotenogenic genes. The identification of
metabolic bottlenecks in the pathway will help to refine strategies for the creation of engineered
plants with specific carotenoid profiles.
Introduction
Plants synthesize a wide variety of natural products via complex
secondary metabolic pathways (Miralpeix et al., 2013; Rischer
et al., 2013). Many of the genes involved in secondary metabolism, transport and storage are poorly characterized, and the
corresponding enzymes are often sequestered in subcellular
compartments to add a further layer of complex regulation
(O’Connor and Maresh, 2006). Therefore, many technical challenges must be overcome before multistep secondary metabolic
pathways can be engineered effectively in heterologous plants.
The endosperm of cereal seeds is a major food staple, but it is
deficient in many vitamins and minerals, including carotenoids
(Zhu et al., 2007, 2008). Some carotenoids are beneficial but
nonessential, for example those acting as antioxidants that help
to prevent diseases such as cancer, whereas pro-vitamin A
carotenoids such as b-carotene are regarded as essential nutrients
because they cannot be synthesized de novo by humans (Bai
et al., 2011; Fraser and Bramley, 2004; Von Lintig and Vogt,
2004). Rice (Oryza sativa) is an important food staple in
developing countries, but carotenoids do not accumulate in the
endosperm, so its consumption as part of a nondiverse diet is
often associated with vitamin A deficiency (Farre et al., 2010;
Underwood and Arthur, 1996).
The carotenoid content and/or composition of staple crops can
be enhanced by the expression of carotenogenic enzymes if
isoprenoid precursors are available, and the carotenoid products
do not accumulate to levels sufficient to oppose the equilibrium
of the reaction. Upstream pathways supplying carotenoid
precursors may therefore influence carotenoid accumulation
n, 2010), and downstream pathways that
(Rodrıguez-Concepcio
metabolize or sequester carotenoids to deplete the carotenoid
pool can also drive the reaction forward (Auldridge et al., 2006;
Campbell et al., 2010; Ohmiya et al., 2006). The naturally
occurring dominant mutation Orange (OR) in cauliflower (Brassica oleracea cv. Botrytis) that induces proplastids and/or
noncoloured plastids to differentiate into chromoplasts has been
transferred to other crops to generate a metabolic sink that
promotes the accumulation of carotenoids (Li and Van Eck, 2007;
Lopez et al., 2008; Lu et al., 2006). Therefore, it should be
possible to influence the rate of carotenoid synthesis in rice by
regulating the supply of precursors and generating a metabolic
sink to store the products (Cazzonelli and Pogson, 2010; Lu and
Li, 2008).
The biosynthesis of carotenoids in rice endosperm is blocked
at the first enzymatic step (phytoene synthase, PSY) which
converts the precursor geranylgeranyl diphosphate (GGPP) into
phytoene. There is also limited flux in the subsequent desaturation reaction, which generates lycopene. Strategies to boost
carotenoid production in rice endosperm therefore involve the
expression of heterologous PSY and the bacterial enzyme CRTI,
which replaces several consecutive reactions that are required
ª 2015 Society for Experimental Biology, Association of Applied Biologists and John Wiley & Sons Ltd
195
196 Chao Bai et al.
to produce lycopene in plants. Although this increases flux
through the early part of the pathway, the expression of these
two enzymes unmasks another rate-limiting step in the supply
of precursors. For example, Golden Rice accumulates bcarotene by the endosperm-specific expression of maize (Zea
mays) PSY (ZmPSY) and Pantoea ananatis CRTI (PaCRTI) under
the control of the rice glutelin promoter, but it does not
accumulate any phytoene, suggesting that the precursor pool is
completely exhausted (Paine et al., 2005; Schaub et al., 2005;
Ye et al., 2000).
To address this challenge, we expressed the heterologous
ZmPSY and PaCRTI enzymes along with Arabidopsis thaliana
1-deoxy-D-xylulose-5-phosphate synthase (AtDXS) specifically in
the endosperm to boost flux through the 2-C-methyl-D-erythritol
4-phosphate (MEP) pathway, which generates carotenoid precursors (Figure 1). A metabolic sink was created in the endosperm
by expressing the A. thaliana ORANGE (AtOR) gene (also under
the control of an endosperm-specific promoter) which has
recently been shown to sequester carotenoids in rice callus (Bai
et al., 2014). It was found that AtDXS combined with ZmPSY1
and PaCRTI significantly enhanced the accumulation of carotenoids in rice endosperm, confirming that the supply of isoprenoid
precursors such as GGPP is a rate-limiting step. The combined
expression of ZmPSY1, PaCRTI and AtOR also boosted carotenoid
accumulation through the creation of a metabolic sink, which
resulted in the up-regulation of several endogenous carotenogenic genes. Our data provide a conceptual and mechanistic basis
to resolve remaining bottlenecks in the carotenoid biosynthesis
pathway in intact plants and will therefore facilitate the development of transgenic plants producing even higher levels of
carotenoids in the endosperm.
Results
The combinatorial expression of carotenogenic genes in
rice endosperm generates diverse genotypes with
different carotenoid profiles
We transformed 7-day-old mature zygotic rice embryos with four
constructs containing unlinked transgenes. Two of the genes
encoded enzymes in the committed carotenoid biosynthesis
pathway (ZmPSY1 and PaCRTI) and a third represented the MEP
pathway (AtDXS). The fourth gene was the selectable marker hpt,
which confers hygromycin resistance. The hpt gene was
expressed constitutively, whereas AtDXS, ZmPSY1 and PaCRTI
were controlled by the endosperm-specific rice RP5 prolamin,
wheat low-molecular-weight glutenin and barley D-hordein promoters, respectively. These experiments generated plants with
different endosperm phenotypes (white, yellow or orange)
depending on the combination of transgenes that were expressed
(Figure 2). The behaviour of the transgenic lines was reproducible
and consistent, that is all lines with the same complement of
transgenes generated near-identical phenotypes.
We also investigated the consequences of sequestering carotenoids in a metabolic sink created by promoting chromoplast
differentiation. To this end, we carried out a second set of
experiments involving four transgenes, namely ZmPSY1, PaCRTI
and the selectable marker hpt as above, but this time also the
AtOR gene (controlled by the wheat low-molecular-weight
glutenin promoter). These experiments also yielded transgenic
rice plants with white, yellow or orange endosperm depending on
the combination of transgenes that were expressed (Figures 2
and 3).
Figure 1 Reconstruction of the carotenoid biosynthesis pathway in rice
endosperm (Bai et al., 2011; Farre et al., 2010). GA3P, glyceraldehyde
3-phosphate; DXP, 1-deoxy-D-xylulose 5-phosphate; DXS, DXP synthase;
AtDXS, Arabidopsis DXS; IPP, isopentenyl diphosphate; IPPI, isopentenyl
diphosphate isomerase; DMAPP, dimethylallyl diphosphate; GGPP,
geranylgeranyl diphosphate; GGPPS, GGPP synthase; ZmPSY1, maize (Zea
mays) phytoene synthase 1; PaCRTI, bacterial (Pantoea ananatis) phytoene
desaturase; PDS, phytoene desaturase; ZISO, f-carotene isomerase; ZDS, fcarotene desaturase; CRTISO, carotenoid isomerase; LYCB, lycopene bcyclase; LYCE, lycopene e-cyclase; CYP97C, carotene e-ring hydroxylase;
HYDB, b-carotene hydroxylase (BCH, CYP97A or CYP97B); ZEP,
zeaxanthin epoxidase; VDE, violaxanthin de-epoxidase; AtOR, Arabidopsis
ORANGE gene. L, seed endosperm expressing ZmPSY1 and PaCRTI; D,
seed endosperm co-expressing AtDXS, ZmPSY1 and PaCRTI; O, seed
endosperm co-expressing AtOR, ZmPSY1 and PaCRTI. Red circles indicate
limiting enzymes numbered sequentially to reflect hierarchical bottlenecks
in the pathway.
The analysis of steady-state mRNA levels showed that plants
with yellow endosperm from both experiments expressed the
ZmPSY1, PaCRTI and hpt genes, whereas those with the orange
endosperm expressed the full complement of transgenes introduced in each experiment (Figure 3). Plants with white endosperm lacked ZmPSY1 and/or PaCRTI expression and were
discarded because the phenotypes were not informative. More
than 20 independent transgenic lines were generated with each
gene combination that produced coloured grains, and from this
pool, we selected three representative lines from each genotype
to allow a more detailed comparison of metabolic profiles. Lines
L1, L2 and L3 expressed ZmPSY1, PaCRTI and hpt only, and the
total carotenoid content was 5.43, 5.51 and 4.61 lg/g dry
weight (DW), respectively. Lines D1, D2 and D3 expressed the
transgenes listed above plus AtDXS, and the total carotenoid
content was 17.79, 14.94 and 31.78 lg/g DW, respectively. Lines
O1, O2 and O3 also expressed the transgenes listed above plus
AtOR, and the total carotenoid content was 11.53, 18.59 and
ª 2015 Society for Experimental Biology, Association of Applied Biologists and John Wiley & Sons Ltd, Plant Biotechnology Journal, 14, 195–205
ZmPSY1 and PaCRTI.
Data are means SD from analysis of three independent seed batches (40 dap) and are expressed as lg/g DW. b/a, the ratio of b,b-carotenoids to b,e-carotenoids; %b-Carotene, the ratio of b-carotene to the total carotenoid
amount (%). Rice endosperm expressing hpt has no detectable carotenoids. O, seed endosperm expressing ZmPSY1, PaCRTI and AtOR; D, seed endosperm expressing ZmPSY1, PaCRTI and AtDXS; L, seed endosperm expressing
52.27
1.40
1.50
50.20
46.99
1.26
0.95
40.73
45.56
1.30
1.42
50.91
25.38
1.08
0.81
39.02
1.06
27.26
b/a
ª 2015 Society for Experimental Biology, Association of Applied Biologists and John Wiley & Sons Ltd, Plant Biotechnology Journal, 14, 195–205
% b-Carotene
1.89 0.28
31.78 5.01
14.94 1.12
1.58 0.12
1.67 0.52
17.79 1.12
25.83 2.37
0.35 0.30
2.36 0.35
18.59 1.64
11.53 2.33
0.47 0.06
1.74 0.06
4.61 0.81
0.50 0.01
5.51 0.16
2.16 0.05
5.43 0.37
Phytoene
1.90 0.00
0.90 0.10
Total Carotenoids
2.09 0.68
9.70 1.60
3.60 0.70
5.00 0.10
9.70 0.80
5.00 1.00
3.20 0.50
0.69 0.39
a-Carotene
0.61 0.04
0.53 0.01
0.70 0.10
1.48 0.10
1.74 0.08
16.61 1.37
a-Cryptoxanthin
0.42 0.08
0.82 0.13
1.49 0.10
1.63 0.16
0.26 0.16
7.50 1.25
8.36 0.44
0.39 0.20
1.73 0.16
10.52 1.64
8.47 1.23
0.58 0.20
0.55 0.07
5.87 1.44
1.17 0.19
0.26 0.09
2.15 0.09
0.26 0.02
1.48 0.02
0.16 0.09
b-Carotene
0.40 0.70
0.29 0.22
0.23 0.14
0.19 0.33
0.15 0.27
0.17 0.13
0.09 0.02
0
0.03
0 0.03
0.17 0.30
Zeaxanthin
b-Cryptoxanthin
Lutein
0.41 0.72
0.33 0.12
0.13 0.22
0.39 0.67
0.41 0.72
0.22 0.15
0.57 0.18
1.13 0.03
0.06 0.01
0
0
0
0
0
0.25 0.03
0
0
0
0.12 0
0.08 0.05
0
0
0
Violaxanthin
0
O1
L3
L2
The expression of the ZmPSY1, PaCRTI, AtDXS and AtOR genes
in the endosperm of transgenic rice plants was analysed by
mRNA blot at 25 days after pollination (dap) as shown in
Figure 3. As anticipated, ZmPSY1 and PaCRTI mRNA accumulated in all three genotypes but not in the hpt-only control,
AtDXS mRNA was detected solely in genotype D, and AtOR
mRNA was detected solely in genotype O (Figure 3).
We also monitored the expression of endogenous carotenogenic genes to determine whether they were regulated in
response to transgene expression. We measured the levels of the
L1
AtOR and AtDXS up-regulate different sets of
endogenous carotenogenic genes
Table 1 Carotenoid content and composition of T3 endosperm at 40 DAP
25.83 lg/g DW, respectively (Table 1). The carotenoid levels in
the orange endosperm of genotype O were therefore 2.1- to
4.7-fold higher than the yellow endosperm of genotype L.
Similarly, the carotenoid levels in the orange endosperm of
genotype D were 2.7- to 5.8-fold higher than genotype L
(Figure 4). The proportion of total carotenoids represented by
b-carotene was 25–39% in genotype L, 47–52% in genotype D
and 40–50% in genotype O, representing average yields of 1.6,
8.3 and 10.8 lg/g DW, respectively (Table 1, Figure S1). The
levels of a-carotene were also higher in genotypes D and O
compared to L, with averages of 6.1, 6.0 and 1.2, lg/g DW,
respectively. Lutein levels were much higher in genotype O
(average 1.0 lg/g DW) than genotypes D (average 0.4 lg/g DW)
or L (average 0.2 lg/g DW) (Table 1, Figure S1).
O2
Figure 3 Transgenic expression analyses in rice endosperm.
Transgenic expression in T3 rice endosperm at 25 dap (25 lg total
RNA per sample). Abbreviations: HPT, seed endosperm expressing hpt;
L1, L2 and L3, seed endosperm expressing ZmPSY1 and PaCRTI; O1,
O2 and O3, seed endosperm expressing ZmPSY1, PaCRTI and AtOR;
D1, D2 and D3, seed endosperm expressing ZmPSY1, PaCRTI and
AtDXS.
0
O3
D1
D2
Figure 2 Phenotype of transgenic rice seeds expressing different
combinations of carotenogenic genes. Two different endosperm
phenotypes were observed. The orange seed endosperm co-expressed
AtDXS, ZmPSY1 and PaCRTI (genotype D), or AtOR, ZmPSY1 and PaCRTI
(genotype O), whereas the yellow seed endosperm (genotype L) coexpressed ZmPSY1 and PaCRTI. Abbreviations: HPT, seed endosperm
expressing hpt; L, seed endosperm expressing ZmPSY1 and PaCRTI; O,
seed endosperm expressing ZmPSY1, PaCRTI and AtOR; D, seed
endosperm expressing ZmPSY1, PaCRTI and AtDXS.
Antheraxanthin
D3
Bottlenecks in carotenoid biosynthesis in rice 197
198 Chao Bai et al.
Figure 4 Carotenoid content and composition of
transgenic rice endosperm. The carotenoid
content and composition of all lines were analysed
by UHPLC. Bars represent different carotenoids in
lines expressing different transgenic
combinations. All carotenoid content data were
averaged for three independent measurements
(T3 mature seeds) at 40 dap. Column names are
defined in the legend to Figure 3.
endogenous phytoene desaturase (OsPDS), lycopene e-cyclase
(OsLYCE), lycopene b-cyclase (OsLYCB), b-carotene hydroxylase
(OsBCH2) and zeaxanthin epoxidase (OsZEP) mRNAs by quantitative real-time PCR at the same time points (Figure 5). All five
endogenous genes were expressed at similar levels in genotype L
and the hpt-only control, indicating the absence of any feedback
regulation in response to the expression of ZmPSY1 and PaCRTI.
However, OsPDS was up-regulated in all three lines of genotype D
and also in one of the lines expressing AtOR (line O3). There was
no induction of OsPDS mRNA in lines O1 and O2, perhaps
reflecting the lower level of AtOR expression in these lines
compared to O3 (Figure 5).
The OsLYCE gene was strongly up-regulated in genotypes D
and O compared to L. The OsLYCB gene was expressed at similar
levels in genotypes L and D and the hpt-only control, but was
induced 1.4- to 1.8-fold in genotype O. OsBCH2 expression was
suppressed in lines D1 and D2 but not in line D3, and the same
gene was up-regulated in all three lines of genotype O (Figure 5).
Finally, OsZEP mRNA levels were up-regulated by up to twofold in
lines O1 and O2 but not in any other lines including O3, which
may again reflect the much higher levels of AtOR mRNA in line
O3 compared to O1 and O2 (Figure 5).
Discussion
Combinatorial transgenic expression results in diverse
phenotypes reflecting the accumulation of different
types and levels of carotenoids in the endosperm
Wild-type rice endosperm does not naturally produce carotenoids, and previous research has shown that significant levels of
carotenoids can only accumulate in this tissue if a heterologous
mini-pathway is imported representing all committed steps prior
to the reaction catalyzed by lycopene b-cyclase. A heterologous
daffodil (Narcissus pseudonarcissus) PSY gene expressed in rice
resulted in the accumulation of phytoene, but no desaturated
products (Burkhardt et al., 1997). Similarly, the expression of
bacterial CRTI (which replaces several consecutive desaturation
and isomerization steps catalyzed by the plant enzymes PDS,
Z-ISO, ZDS and CRTISO) alone was insufficient to relieve the
bottleneck and no carotenoids accumulated in the endosperm
due to the absence of PSY activity (Chao Bai, Changfu Zhu,
Teresa Capell, Paul Christou.). Therefore, carotenoid accumulation in rice requires at least two enzymes: PSY and CRTI. This
approach was demonstrated during the development of Golden
Rice, which expressed daffodil PSY and PaCRTI to produce
1.6 lg/g DW of carotenoids in the endosperm (Ye et al., 2000).
The yield was later improved to 37 lg/g DW by replacing the
daffodil enzyme with the more active ZmPSY1 (Paine et al.,
2005).
Both the Golden Rice and Golden Rice II lines accumulate
predominantly b-carotene at the expense of other carotenoid
intermediates earlier in the pathway, and they are notable for the
almost complete absence of phytoene, the immediate product of
PSY1 (Paine et al., 2005; Ye et al., 2000). These results suggest
that there is an additional bottleneck upstream of the first
committed pathway step, which means that the pathway
precursors are almost entirely depleted. Furthermore, the buildup of b-carotene is likely to affect the equilibrium in the cell and
inhibit further conversion, suggesting that b-carotene levels could
be boosted further by increasing the precursor supply and
removing the product by transferring it to a metabolic sink. In
an earlier study, we therefore used our callus platform for the
rapid functional characterization of genes to test the expression
of AtDXS (to increase flux through the MEP pathway and
therefore increase the availability of precursors for carotenoid
synthesis) and AtOR, which induces the differentiation of
chromoplasts and thus promotes the accumulation of carotenoids
in the plastid (Bai et al., 2014). Surprisingly, we found that
chromoplast differentiation could be induced directly not only by
the wild-type AtOR gene but also by increasing the supply of
precursors in the absence of AtOR (Bai et al., 2014).
Although the callus system is a useful in vitro platform, it is
unlikely to represent accurately the intracellular milieu present
in differentiated endosperm tissue. We therefore used our
established combinatorial expression platform that allows the
recovery of diverse metabolic libraries based on random
combinations of input transgenes (Zhu et al., 2008) to produce
transgenic lines expressing ZmPSY1 and PaCRTI (genotype L)
complemented with either AtDXS (genotype D) or AtOR
(genotype O). All the genotype L plants produced yellow
grains reflecting the accumulation of carotenoids in the
endosperm, and likewise the genotype D and genotype O
plants produced orange grains. Other transgenic lines were
recovered expressing either AtDXS or AtOR in the absence of
ZmPSY1 and PaCRTI, but because these plants lacked essential
steps in the early carotenoid synthesis pathway, they did not
accumulate any carotenoids and the grains were white. Lines
expressing only the selectable marker gene hpt were also
recovered, and these were used as controls together with the
lines expressing only AtDXS or AtOR.
ª 2015 Society for Experimental Biology, Association of Applied Biologists and John Wiley & Sons Ltd, Plant Biotechnology Journal, 14, 195–205
Bottlenecks in carotenoid biosynthesis in rice 199
Figure 5 Expression of endogenous carotenogenic genes in transgenic rice endosperm. Relative transcript levels of endogenous OsPDS, OsLYCE, OsLYCB,
OsBCH2 and OsZEP genes in T3 rice endosperm at 25 dap expressing different transgenic combinations. Values are means SD of three quantitative
real-time PCR replicates. The expression for each gene was normalized against actin mRNA. Abbreviations: PDS, phytoene desaturase; LYCE, lycopene ecyclase; LYCB, lycopene b-cyclase; BCH, b-carotene hydroxylase; ZEP, zeaxanthin epoxidase. Line names are defined in the legend to Figure 3.
Increasing the precursor supply enhances carotenoid
accumulation in rice endosperm and modulates
endogenous carotenogenic genes
Carotenoids are synthesized primarily from precursors derived from
the MEP pathway shown in Figure 1 (Eisenreich et al., 2001; Farre
n, 2010; Rodrıguez-Concepcio
n
et al., 2010; Rodrıguez-Concepcio
and Boronat, 2002). This pathway uses pyruvate and glyceraldehyde 3-phosphate to produce 1-deoxy-D-xylulose-5-phosphate
(DXP) in a reaction catalyzed by DXP synthase (DXS). DXS is the
rate-limiting step in the formation of plastid-derived isoprenoids
(Enfissi et al., 2005; Est
evez et al., 2001; Morris et al., 2006;
n, 2010). Therefore, the fruit-specific overRodrıguez-Concepcio
expression of Escherichia coli DXS in tomatoes increased the total
carotenoid content by up to 1.6-fold although most of this
accumulated as phytoene and the end products lycopene and
b-carotene were not boosted at all, suggesting that phytoene
desaturation was a new rate-limiting step in the transgenic plants
(Enfissi et al., 2005). Similar results were achieved in transgenic
potatoes expressing E. coli DXS (Morris et al., 2006).
To determine whether DXS can enhance the isoprenoid
precursor pool in rice endosperm, we compared transgenic rice
lines co-expressing AtDXS, ZmPSY1 and PaCRTI (genotype D)
with those co-expressing ZmPSY1 and PaCRTI (genotype L). The
ª 2015 Society for Experimental Biology, Association of Applied Biologists and John Wiley & Sons Ltd, Plant Biotechnology Journal, 14, 195–205
200 Chao Bai et al.
endosperm in genotype D was orange compared to the yellow
endosperm of genotype L, indicating a qualitative and/or
quantitative difference in carotenoid levels. Accordingly, we
found that the three lines of genotype D accumulated 2.7- to
5.8-fold more total carotenoids than genotype L (Table 1). This
was a greater increase than achieved with the same gene
combination in callus (Bai et al., 2014), which confirmed that
although the callus system is useful for rapid functional characterization, it cannot represent the more complex environment of
the fully differentiated endosperm. In contrast to the tomato and
potato lines described above (Enfissi et al., 2005; Morris et al.,
2006), the levels of phytoene in rice genotypes D and L were
similar, with averages of 1.5 and 1.7 lg/g DW, respectively
(Table 1); thus, the higher carotenoid levels in genotype D mainly
reflected increases in the levels of b-carotene, a-cryptoxanthin
and a-carotene (Table 1). This correlates well with the observed
up-regulation of endogenous phytoene desaturase (OsPDS) and
lycopene e-cyclase (OsLYCE) in these lines (Figure 5) because the
combination of heterologous DXS and endogenous and heterologous PDS would increase flux through the entire pathway,
whereas LYCE would specifically boost the production of
a-carotene and a-cryptoxanthin (without the loss of b-carotene
given the overall increase in flux caused by PDS activity). In
contrast, the unaltered LYCB and BCH2 activity may give rise to a
new bottleneck in response to the increased flux, thus resulting in
the accumulation of b-cryptoxanthin prior to its conversion into
beta-carotene. The total carotenoid content of endosperm
expressing ZmPSY1, PaCRTI and AtDXS was at least threefold
higher than that expressing only ZmPSY1 and PaCRTI (Table 1,
Figure 4), indicating the supply of isoprenoid precursors such as
GGPP derived from the MEP pathway is critical for maximizing
carotenoid accumulation in rice endosperm. In the callus system,
expression of the same three genes resulted mainly in the
accumulation of b-carotene, a-carotene and phytoene, again
reflecting differences between callus and endosperm in terms of
the relative levels of precursors, enzymes and intermediates (Bai
et al., 2014).
Creating a metabolic sink enhances carotenoid
accumulation in rice endosperm and modulates
endogenous carotenogenic genes
The cauliflower Or gene is the only known gene that acts as a
bona fide molecular switch to trigger the differentiation of
noncoloured plastids into chromoplasts (Giuliano and Diretto,
2007; Lu et al., 2006). It was discovered as a spontaneous
dominant mutation in cauliflower, which caused minimally
pigmented tissues such as the edible curd to become orange
due to the accumulation of high levels of b-carotene (Li et al.,
2001; Lu et al., 2006). The expression of cauliflower Or in
transgenic potato tubers caused the ectopic induction of chromoplasts, allowing the tubers to accumulate substantial amounts
of b-carotene, violaxanthin, lutein, phytoene, phytofluene and
f-carotene (Lopez et al., 2008). There was no indication in these
experiments that the cauliflower OR induced the expression of
carotenogenic genes (Li et al., 2001, 2006; Lu et al., 2006).
However, the overexpression of a sweet potato OR orthologue
was recently shown to induce carotenogenic gene expression in
transgenic sweet potato callus, suggesting that OR might
promote carotenoid accumulation to the extent that chromoplast
differentiation is triggered by the presence of excess carotenoids
rather than the direct activity of the OR protein (Kim et al., 2013).
Recently, we demonstrated that the overexpression of the
wild-type AtOR gene induced chromoplast formation in rice
callus and enhanced carotenoid levels by forming a metabolic
sink (Bai et al., 2014). However, the results from our analysis of
the genotype L and genotype O transgenic plants described
herein present a more complex picture, in which AtOR expression may influence carotenoid levels both directly and indirectly.
The genotype O transgenic lines accumulated 2.1- to 4.7-fold
more total carotenoids in the endosperm than genotype L, and
most of this increase was accounted for by the 5.2-fold increase
in the levels of b-carotene, b-cryptoxanthin and a-carotene, as
well as a 2.1- to 6.6-fold increase in the levels of lutein
(Table 1). This correlated well with the observed induction of
the endogenous OsLYCE, OsLYCB and OsBCH2 genes and
confirmed that AtOR promotes the modulation of endogenous
carotenogenic gene expression in rice (Figure 5). The role of
AtOR in chromoplast differentiation appears to be more
complex than initially envisaged, because we noted in earlier
experiments the formation of chromoplasts in rice callus of
genotype O (expressing AtOR and the carotenoid mini-pathway)
and genotype D (expressing AtDXS and the carotenoid minipathway but not AtOR), but we did not find any evidence for
chromoplast differentiation in the endosperm of plants expressing AtOR in the absence of ZmPSY1 and PaCRTI. Transmission
electron microscopy revealed a few electron-dense plastoglobuli
(which may contain pigment) inside some plastids of 40 dap
endosperm in the genotypes O and D, and various small
plastoglobuli inside some plastids were visible in lines L, whereas
no plastoglobuli were visible in lines HPT (expressing only the
selectable marker gene hpt) (Figure 6). This suggests that
chromoplast differentiation is primarily triggered by carotenoid
accumulation above a certain threshold and that the presence of
the Orange protein may augment or potentiate this process but
is not sufficient without other drivers of carotenoid accumulation (Bai et al., 2014; Maass et al., 2009). Moreover, perturbations in carotenoid composition of Psy-1 transgenic tomato
induced plastid differentiation during fruit development (Fraser
et al., 2007).
AtDXS and AtOR regulate carotenoid biosynthesis in rice
endosperm through distinct mechanisms
Our experiments revealed that AtDXS and AtOR regulate different
sets of endogenous carotenogenic genes, suggesting they boost
carotenoid accumulation via distinct mechanisms. In genotype O
lines expressing AtOR, we observed the up-regulation of OsLYCE,
OsLYCB, OsBCH2 and OsZEP, and only OsPDS was slightly
affected (Figure 5). In contrast, in genotype D lines expressing
AtDXS, we observed the up-regulation of OsPDS and OsLYCE but
not the other three genes. Both AtDXS and AtOR thus boosted
the accumulation of total carotenoids in rice endosperm by at
least twofold when expressed in concert with ZmPSY1 and
PaCRTI (Table 1). In both cases, this predominantly reflected the
presence of higher levels of b-carotene and a-carotene (Table 1
and Figure 4). The level of endogenous OsLYCB increased in lines
expressing AtOR but not AtDXS (Figure 5), suggesting endogenous OsLYCB is sufficient for enhanced b-carotene synthesis in
lines overexpressing DXS as might be expected given the greater
flux throughout the entire pathway. The level of endogenous
OsLYCE mRNA was significantly up-regulated in lines expressing
either AtDXS or AtOR (Figure 5), which explains the fivefold
increase in the accumulation of a-carotene (Table 1).
ª 2015 Society for Experimental Biology, Association of Applied Biologists and John Wiley & Sons Ltd, Plant Biotechnology Journal, 14, 195–205
Bottlenecks in carotenoid biosynthesis in rice 201
(a)
(b)
(c)
(d)
Figure 6 Transmission electron micrographs of
transgenic rice endosperm expressing ZmPSY1,
PaCRTI and AtOR (a), ZmPSY1, PaCRTI and AtDXS
(b), ZmPSY1 and PaCRTI (c), HPT (d). Arrows
indicate plastoglobuli inside plastids (pl). Scale bar
0.3 lm (a, b and c) and 1 lm (d).
Zeaxanthin and lutein levels were higher in the lines expressing AtOR, which was consistent with the induction of endogenous OsBCH2 expression (Table 1 and Figure 5). However,
similar zeaxanthin and lutein levels were observed in the lines
expressing AtDXS, in which endogenous OsBCH2 expression
remained at normal levels (Table 1 and Figure 5). This argues
that the flux increased in both lines, in one case due to the
greater abundance of GGPP and in the other by the removal of
a bottleneck at the hydroxylation step, reflecting the stronger
expression of OsBCH2.
Phytoene levels were lower in genotype O (average 1.1 lg/g
DW) than genotype D (average 1.5 lg/g DW) or genotype L
(average 1.7 lg/g DW) (Table 1). Nevertheless, phytoene accumulated in all nine lines, suggesting that CRTI is not efficient
enough to convert all phytoene into downstream carotenoids,
which was not the case for Golden Rice (Paine et al., 2005; Ye
et al., 2000). Phytoene was also detected in the plastoglobules
of transgenic tomato fruit chromoplasts, whereas the corresponding enzymes are localized in the membranes (Nogueira
et al., 2013). It is possible that phytoene and CRTI may also be
compartmentalized separately in transgenic rice endosperm,
resulting in residual phytoene accumulation, highlighting the
important role of compartmentalization in the control of
endogenous and introduced metabolic pathways.
A variety of factors determine the qualitative and
quantitative profile of carotenoids in rice endosperm
The discordance between the endosperm carotenoid profiles in
our rice lines and the previously described Golden Rice and
Golden Rice II lines is noteworthy. One factor that may play a
significant role is the cultivar, which was Kaybonnet in the case
of Golden Rice II and EYI105 in our experiments, because the
level of endogenous MEP and carotenoid pathway gene
expression may differ (Paine et al., 2005). This may explain
why the endosperm carotenoid composition of our lines was
25–39% b-carotene and 9–11% a-cryptoxanthin (Table 1),
whereas in Golden Rice II, the equivalent values were 75–84%
b-carotene and a-cryptoxanthin was not detectable (Paine et al.,
2005). This suggests cultivar-specific differences in the endogenous b-carotene and a-carotene hydroxylase activities, consistent with the natural variation in the expression of
carotenogenic genes not only in different varieties of rice but
also in other cereals such as maize (Harjes et al., 2008) and
sorghum (Kean et al., 2007).
Other differences may reflect the control of transgenic
expression. We used the wheat low-molecular-weight glutenin
promoter to control ZmPSY1 and the barley D-hordein promoter
to control PaCRTI, whereas in Golden Rice II, both transgenes
ª 2015 Society for Experimental Biology, Association of Applied Biologists and John Wiley & Sons Ltd, Plant Biotechnology Journal, 14, 195–205
202 Chao Bai et al.
were controlled by the rice glutelin promoter (Paine et al., 2005).
We have previously noted that different endosperm-specific
promoters affect the mRNA and protein levels of the corresponding enzymes and also the level of any metabolites in the
engineered pathway (Naqvi et al., unpublished).
The rate of carotenoid degradation or turnover inversely
determines carotenoid content (Nisar et al., 2015). A family of
carotenoid cleavage dioxygenases (CCDs) catabolizes specific
enzymatic turnover of carotenoids into apocarotenoids in maize,
rice and sorghum (Vallabhaneni et al., 2010). The expression of
carotenoid cleavage dioxygenase 1 was inversely correlated with
the accumulation of carotenoids in maize endosperm (da Silva
Messias et al., 2014; Vallabhaneni et al., 2010). The activities of
various carotenoid cleavage dioxygenases might be different in
Kaybonnet in the case of Golden Rice II and EYI105 in our
experiments. Our results clearly demonstrate a mechanistic basis
for differences in carotenoid profiles between plants with
different transgenic complements and genetic backgrounds,
and also significant differences between the differentiated
endosperm of intact plants and the in vitro screening system
based on rice callus (Bai et al., 2014).
Conclusions
We have investigated factors that limit the accumulation of
carotenoids in rice endosperm, specifically focusing on the
precursor pool and the impact of a metabolic sink. The metabolic
comparison of endosperm from genotype L plants (ZmPSY1 and
PaCRTI) and genotype D plants (ZmPSY1, PaCRTI and AtDXS)
confirmed that the supply of isoprenoid precursors from the MEP
pathway is a key bottleneck, which limits flux through the entire
pathway. The metabolic comparison of genotype L plants and
genotype O plants (ZmPSY1, PaCRTI and AtOR) clearly demonstrated that the overexpression of AtOR can boost carotenoid
accumulation in rice endosperm by creating a metabolic sink that
drives the equilibrium of the pathway towards completion, and
also by the up-regulation of endogenous carotenogenic genes.
The twofold increase in total carotenoids achieved in genotypes D
and O predominantly reflected the accumulation of provitamin A
carotenoids to at least fivefold normal levels, based on the
combined mechanisms of increased precursor supply, the creation of a metabolic sink and the induction of endogenous gene
expression in the committed part of the carotenoid pathway.
However, AtDXS and AtOR affected overlapping sets of endogenous carotenogenic genes through different mechanisms.
AtDXS induced the expression of OsPDS and OsLYCE, whereas
AtOR induced all the endogenous carotenogenic genes we
tested, with the exception of OsPDS. These experiments offer
insight into the bottlenecks in the carotenoid pathway in rice
endosperm and will allow the development of more targeted
strategies for the creation of engineered plants with particular
carotenoid profiles.
Gene cloning and vector construction
The AtDXS and AtOR cDNAs were cloned directly from A. thaliana leaf mRNA by RT-PCR using primers designed according to
sequences in GenBank (GenBank accession number: NM 203246
and U27099.1). The cDNAs were transferred to vector pGEM-T
Easy (Promega, Madison, WI, USA), and the recombinant vectors
were digested with EcoRI to release the cDNAs. AtDXS was
inserted into vector pRP5 (Su et al., 2001) containing the
endosperm-specific rice prolamin promoter and the ADPGPP
terminator, and AtOR was inserted into vector p326 (Stoger
et al., 1999) containing the endosperm-specific wheat lowmolecular-weight glutenin gene promoter and nos terminator.
Both vectors were linearized with EcoRI to accept the inserts.
The maize PSY1 cDNA was cloned from maize inbred line B73
endosperm by RT-PCR using forward primer 50 -AGG ATC CAT
GGC CAT CAT ACT CGT ACG AG-30 incorporating a BamHI site
(underlined) and reverse primer 50 -AGA ATT CTA GGT CTG GCC
ATT TCT CAA TG-30 incorporating an EcoRI site (underlined). The
primers were designed based on sequences in GenBank (GenBank accession number: AY324431). The product was transferred
to vector pGEM-T Easy (Promega) to generate pGEM-ZmPSY1 for
sequencing and then inserted into vector p326 as above (Stoger
et al., 1999).
The Pantoea ananatis (formerly Erwinia uredovora) phytoene
desaturase gene (CRTI) was fused in frame with the transit
peptide signal from the Phaseolus vulgaris small subunit of
ribulose bisphosphate carboxylase (Schreier et al., 1985) in
plasmid pYPIET4 (Misawa et al., 1994) and amplified by PCR
using forward primer 50 -ATC TAG AAT GGC TTC TAT GAT ATC
CTC TTC-30 incorporating an XbaI site (underlined) and reverse
primer 50 -AGA ATT CTC AAA TCA GAT CCT CCA GCA TCA-30
incorporating an EcoRI site (underlined). The primers were
designed based on sequences in GenBank (GenBank accession
number: D90087). The product was transferred to vector pGEM-T
Easy to produce pGEM-PaCRTI for sequencing and then inserted
into pHorp-P containing the endosperm-specific barley D-hordein
promoter and the rice ADPGPP terminator (Sorensen et al.,
1996).
Transformation of rice plants
Experimental procedures
Seven-day-old mature zygotic rice embryos were transferred to
MS osmoticum medium (4.4 g/L Murashige-Skoog salts supplemented with 0.3 g/L casein hydrolysate, 0.5 g/L proline, 72.8 g/L
mannitol and 30 g/L sucrose) 4 h before transformation and then
bombarded with 10 mg gold particles coated with the carotenogenic constructs and the selectable maker hpt at a 3 : 3 : 1 or
3 : 3 : 3 : 1 ratio as appropriate (Christou, 1997; Christou et al.,
1991). The embryos were returned to osmoticum medium for
12 h before selection on MS medium (as above but without
mannitol) supplemented with 30 mg/L hygromycin and 2.5 mg/L
2,4-dichlorophenoxyacetic acid in the dark for 2–3 weeks (Farr
e
et al., 2012). Transgenic plantlets were regenerated and hardened off in soil.
Plant material
mRNA blot analysis
Wild-type rice (Oryza sativa L. cv. EYI105) and transgenic rice
plants were grown in the greenhouse and growth chamber at
28/20 °C day/night temperature with a 10-h photoperiod and
60–90% relative humidity for the first 50 days, followed by
maintenance at 21/18 °C day/night temperature with a 16-h
photoperiod thereafter.
Total RNA was extracted from rice endosperm 25 dap. We
separated 25 lg denatured total RNA by 1.2% (w/v) agaroseformaldehyde gel electrophoresis in 19 MOPS buffer and
transferred the fractionated RNA to a membrane by capillary
blotting (Sambrook et al., 1989). The membrane was probed
with digoxigenin-labelled partial cDNAs prepared using the
ª 2015 Society for Experimental Biology, Association of Applied Biologists and John Wiley & Sons Ltd, Plant Biotechnology Journal, 14, 195–205
Bottlenecks in carotenoid biosynthesis in rice 203
PCR-DIG Probe Synthesis Kit (Roche, Mannheim, Germany),
with hybridization carried out at 50 °C overnight using DIG
Easy Hyb (Roche). The membrane was washed twice for 5 min
in 29 SSC, 0.1% SDS at room temperature, twice for 20 min
in 0.29 SSC, 0.1% SDS at 68 °C and then twice for 10 min in
0.19 SSC, 0.1% SDS at 68 °C. After immunological detection
with anti-DIG-AP (Fab-Fragments Diagnostics GmbH, Roche,
Mannheim, Germany), chemiluminescence generated by disodium [3-(1-chloro-30 -methoxy-spiro[adamantane-4,40 -dioxetane]30 -yl)phenyl] phosphate (CSPD) (Roche) was detected on Kodak
BioMax light film (Sigma-Aldrich, St. Louis, MO) according to
the manufacturer’s instructions. The forward and reverse
primers for each transgene used for probe synthesis are shown
in Table S1.
Carotenoid extraction and quantification
Carotenoids were extracted in darkness from 50 mg freeze-dried
endosperm (40 dap) with 50/50 (vol/vol) tetrahydrofuran and
methanol at 60 °C for 20 min. The mixture was filtered and the
residue re-extracted in acetone. Chromatography was carried out
using a Waters ACQUITY UPLCTM system (Waters, Milford, MA)
comprising an ACQUITY UPLCTM binary solvent manager and an
ACQUITY UPLCTM sample manager, coupled to a photodiode
array (PDA) 2996 detector. This was linked to an AcquityTM TQD
tandem-quadrupole MS equipped with a Z-spray electrospray
interface (Manchester, UK). MassLynxTM software version 4.1
(Waters) was used for data acquisition and processing. Compounds were separated with an ACQUITY UPLCTM BEH C18
column (1.7 lm; 1 9 150 mm) (Waters) and a gradient system
with the mobile phase consisting of solvent A (7 : 3 v/v
acetone : methanol) and solvent B (100% water). The linear
gradient was as follows: 0–1.0 min, 25% B, 0.5 mL/min (isocratic); 1.0–10.0 min, 4.9% B 0.5 mL/min (linear gradient); 10.0–
11.4 min, 0% B, 0.7 mL/min (linear gradient); 11.4–19.2 min,
0% B, 0.7 mL/min (isocratic); 19.2–20.0 min, 25% B, 0.5 mL/min
(linear gradient); and 20.0–22.0 min, 25% B, 0.5 mL/min
(isocratic). The injection needle was washed with acetone : methanol (7 : 3 v/v) and subsequently with 2-propanol between
injections. The injection volume was 5 lL and the column was
kept at 32 °C while the temperature in the sample manager was
maintained at 25 °C. The average maximum pressure in the
chromatographic system was 15 000 psi (Delpino-Rius et al.,
2014).
Carotenoids were identified using reference compounds as
standards and by their spectral properties. Lutein, b-cryptoxanthin and b-carotene were purchased from Sigma (St Louis,
MO); zeaxanthin from Fluka (Buchs, Switzerland); and phytoene, a-carotene, violaxanthin and antheraxanthin from CaroteNature (Lupsingen, Switzerland). The hydroxylated carotenoids
were further characterized by mass spectroscopy. Optical
absorbance maxima and masses were for a-cryptoxanthin
420, 445, 475 nm with m/z 552; for b-cryptoxanthin 425,
450, 475 nm with m/z 552; for lutein 422, 445, 475 nm with
m/z 568; and for zeaxanthin 425, 450, 476 nm with m/z 568.
Quantification was carried out with calibration curves of the
standards. In the case of a-cyptoxanthin, the calibration curve
for a-carotene was used as the latter carotenoid exhibits very
similar absorbance properties. Total carotenoid content in
extracts was calculated spectrophotometrically using the absorbance at 450 nm and the average extinction coefficient A1%
1cm of
2332 as lutein and zeaxanthin represented the majority of the
carotenoids.
Quantitative real-time PCR
Real-time PCR was used to amplify RNA isolated from rice
endosperm (25 dap) on a Bio-Rad CFX96TM system using a 25ll mixture containing 10 ng of synthesized cDNA, 1 3 iQ SYBR
green supermix (Bio-Rad, Hercules, CA) and 0.2 lM forward and
reverse primers for the target genes. To calculate relative
expression levels, serial dilutions (0.2–125 ng) were used to
produce standard curves for each gene. PCRs were carried out
in triplicate using 96-well optical reaction plates, comprising a
heating step for 3 min at 95 °C, followed by 40 cycles of 95 °C
for 15 s, 58.5 °C for 1 min and 72 °C for 20 s. Amplification
specificity was confirmed by melt curve analysis on the final PCR
products in the temperature range 50–90 °C with fluorescence
acquired after each 0.5 °C increment. The fluorescence
threshold value and gene expression data were calculated using
the CFX96 system software Bio-Rad, Hercules, CA. Values
represent the mean of three real-time PCR replicates SD. The
forward and reverse primers for each transgene are shown in
Table S2.
Transmission electron microscopy
Rice seed sections (0.5 9 2.0 mm) were fixed in 2.5% v/v
glutaraldehyde and 2.0% v/v paraformaldehyde in 0.1 M sodium
phosphate buffer (pH 7.2) overnight at 4 °C. The sections were
washed three times in 0.1 M sodium phosphate buffer (pH 7.2),
and post-fixed in 1% w/v osmium tetroxide in 0.1 M sodium
phosphate buffer (pH 7.2) for 1 h. They were then washed three
times in re-distilled water and dehydrated in an alcohol series
(30–100%) before embedding in epoxy resin Araldite Embed
812 (Epon-812) from Aname Electron Microscopy Sciences,
Madrid, Spain (URL: www.aname.es) and polymerizing at
60 °C. Ultra-thin sections (80–90 nm) were prepared with a
diamond knife using a Reichter Jung Ultramicrotome Ultracut E
(Scotia), mounted on SPI-ChemTM Formvar/carbon-coated copper grids, and stained with uranyl acetate and Reynold’s lead
citrate prior to examination using an EM 910 transmission
electron microscope (Carl Zeiss, Jena, Germany).
Acknowledgements
This study was supported by the Ministerio de Economia y
Competitividad, Spain (BIO2011-22525 and PIM2010PKB-00746
CAROMAIZE), European Research Council Advanced Grant (BIOFORCE) to PC. C.B. is the recipient of a PhD fellowship from the
Universitat de Lleida, Spain.
References
Auldridge, M.E., Block, A., Vogel, J.T., Dabney-Smith, C., Mila, I., Bouzayen,
M., Magallanes-Lundback, M., DellaPenna, D., McCarty, D.R. and Klee, H.J.
(2006) Characterization of three members of the Arabidopsis carotenoid
cleavage dioxygenase family demonstrates the divergent roles of this
multifunctional enzyme family. Plant J. 45, 982–993.
Bai, C., Twyman, R.M., Farre, G., Sanahuja, G., Christou, P., Capell, T. and Zhu,
C. (2011) A golden era-pro-vitamin A enhancement in diverse crops. In Vitro
Cell. Dev. Biol. Plant, 47, 205–221.
Bai, C., Rivera, S.M., Medina, V., Alves, R., Vilaprinyo, E., Sorribas, A., Canela,
R., Capell, T., Sandmann, G., Christou, P. and Zhu, C. (2014) An in vitro
system for the rapid functional characterization of genes involved in
carotenoid biosynthesis and accumulation. Plant J. 77, 464–475.
Burkhardt, P.K., Beyer, P., Wunn, J., Kloti, A., Armstrong, G.A., Schledz, M.,
von, L.J. and Potrykus, I. (1997) Transgenic rice (Oryza sativa) endosperm
ª 2015 Society for Experimental Biology, Association of Applied Biologists and John Wiley & Sons Ltd, Plant Biotechnology Journal, 14, 195–205
204 Chao Bai et al.
expressing daffodil (Narcissus pseudonarcissus) phytoene synthase
accumulates phytoene, a key intermediate of provitamin A biosynthesis.
Plant J. 11, 1071–1078.
Campbell, R., Ducreux, L.J., Morris, W.L., Morris, J.A., Suttle, J.C., Ramsay, G.,
Bryan, G.J., Hedley, P.E. and Taylor, M.A. (2010) The metabolic and
developmental roles of carotenoid cleavage dioxygenase 4 form potato.
Plant Physiol. 154, 656–664.
Cazzonelli, C.I. and Pogson, B.J. (2010) Source to sink: regulation of
carotenoids biosynthesis in plants. Trends Plant Sci. 15, 266–274.
Christou, P. (1997) Rice transformation by bombardment. Plant Mol. Biol. 35,
197–203.
Christou, P., Ford, T. and Kofron, M. (1991) Production of transgenic rice (Oryza
sativa L.) plants from agronomically important indica and japonica varieties
via electric discharge particle acceleration of exogenous DNA into immature
zygotic embryos. Nat. Biotechnol. 9, 957–962.
, F., Balcells, M. and CanelaDelpino-Rius, A., Eras, J., Marsol-Vall, A., Vilaro
Garayoa, R. (2014) Ultra performance liquid chromatography analysis to
study the changes in the carotenoid profile of commercial monovarietal fruit
juices. J. Chromatogr. A, 1331, 90–99.
Eisenreich, W., Rohdich, F. and Bacher, A. (2001) Deoxyxylulose phosphate
pathway to terpenoids. Trends Plant Sci. 6, 78–84.
Enfissi, E.M.A., Fraser, P.D., Lois, L.M., Boronat, A., Schuch, W. and Bramley, P.
(2005) Metabolic engineering of the mevalonate and non-mevalonate
isopentenyl diphosphate-forming pathways for the production of healthpromoting isoprenoids in tomato. Plant Biotechnol. J. 3, 17–27.
Est
evez, J.M., Cantero, A., Reindl, A., Reichler, S. and Leon, P. (2001) 1-DeoxyD-xylulose-5-phosphate synthase, a limiting enzyme for plastidic isoprenoid
biosynthesis in plants. J. Biol. Chem. 276, 22901–22909.
Farre, G., Sanahuja, G., Naqvi, S., Bai, C., Capell, T., Zhu, C.F. and Christou, P.
(2010) Travel advice on the road to carotenoids in plants. Plant Sci. 179, 28–
48.
pez, U., Rivera, S.M., Canela,
Farr
e, G., Naqvi, S., Sanahuja, G., Bai, C., Zorrilla-Lo
R., Sandman, G., Twyman, R.M., Capell, T., Zhu, C. and Christou, P. (2012)
Combinatorial genetic transformation of cereals and the creation of metabolic
libraries for the carotenoid pathway. Methods Mol. Biol. 847, 419–435.
Fraser, P.D. and Bramley, P.M. (2004) The biosynthesis and nutritional uses of
carotenoids. Prog. Lipid Res. 43, 228–265.
Fraser, P.D., Enfissi, E.M.A., Halket, J.M., Truesdale, M.R., Yu, D., Gerrish, C.
and Bramley, P.M. (2007) Manipulation of phytoene levels in tomato fruit:
effects on isoprenoids, plastids, and intermediary metabolism. Plant Cell, 19,
3194–3211.
Giuliano, G. and Diretto, G. (2007) Of chromoplasts and chaperones. Trends
Plant Sci. 12, 529–531.
Harjes, C.E., Rocheford, T.R., Bai, L., Brutnell, T.P., Kandianis, C.B., Sowinski,
S.G., Stapleton, A.E., Vallabhaneni, R., Williams, M., Wurtzel, E.T., Yan, J.
and Buckler, E.S. (2008) Natural genetic variation in Lycopene Epsilon Cyclase
tapped for maize biofortification. Science, 319, 330–333.
Kean, E.G., Ejeta, G., Hamaker, B. and Ferruzzi, M.G. (2007) Characterization
of carotenoid pigments in mature and developing kernels of select yellowendosperm sorghum varieties. J. Agric. Food Chem. 55, 2619–2626.
Kim, S.H., Ahn, Y.O., Ahn, M.J., Jeong, J.C., Lee, H.S. and Kwak, S.S. (2013)
Cloning and characterization of an Orange gene that increases carotenoid
accumulation and salt stress tolerance in transgenic sweetpotato cultures.
Plant Physiol. Biochem. 70, 445–454.
Li, L. and Van Eck, J. (2007) Metabolic engineering of carotenoids accumulation
by creating a metabolic sink. Transgenic Res. 16, 581–585.
Li, L., Paolillo, D.J., Parthasarathy, M.V., DiMuzio, E.M. and Garvin, D.F. (2001)
A novel gene mutation that confers abnormal patterns of beta-carotene
accumulation in cauliflower (Brassica oleracea var. botrytis). Plant J. 26, 59–
67.
Li, L., Lu, S., Cosman, K.M., Earle, E.D., Garvin, D.F. and O’Neill, J. (2006)
b-carotene accumulation induced by the cauliflower Or gene is not due to an
increased capacity of biosynthesis. Phytochemistry, 67, 1177–1184.
Lopez, A.B., Van Eck, J., Conlin, B.J., Paolillo, D.J., O’Neill, J. and Li, L. (2008)
Effect of the cauliflower Or transgene on carotenoids accumulation and
chromoplast formation in transgenic potato tubers. J. Exp. Bot. 59, 213–223.
Lu, S. and Li, L. (2008) Carotenoid metabolism: biosynthesis, regulation, and
beyond. J. Integr. Plant Biol. 50, 778–785.
Lu, S., Van Eck, J., Zhou, X.J., Lopez, A.B., O’Halloran, D.M., Cosman, K.M.,
Conlin, B.J., Paolillo, D.J., Garvin, D.F., Vrebalov, J., Kochian, L.V., Kupper, H.,
Earle, E.D., Cao, J. and Li, L. (2006) The cauliflower Or gene encodes a DnaJ
cysteine-rich domain-containing protein that mediates high-levels of
b-carotene accumulation. Plant Cell, 18, 3594–3605.
Maass, D., Arango, J., Wust, F., Beyer, P. and Welsch, R. (2009) Carotenoid
crystal formation in Arabidopsis and carrot roots caused by increased
phytoene synthase protein levels. PLoS ONE, 4, e6373.
Miralpeix, B., Rischer, H., H€akkinen, S.T., Ritala, A., Sepp€anen-Laakso, T.,
Oksman-Caldentey, K.M., Capell, T. and Christou, P. (2013) Metabolic
engineering of plant secondary products: which way forward? Curr. Pharm.
Des. 19, 5622–5639.
Misawa, N., Yamano, S., Linden, H., de Felipe, M.R., Lucas, M., Ikenaga, H. and
Sandmann, G. (1994) Functional expression of the Erwinia uredovora
carotenoid biosynthesis gene crtI in transgenic plants showing an increase
of b-carotene biosynthesis activity and resistance to the bleaching herbicide
norflurazon. Plant J. 4, 833–840.
Morris, W.L., Ducreux, L.J.M., Hedden, P., Millam, S. and Taylor, M.A. (2006)
Overexpression of a bacterial 1-deoxy-D-xylulose 5-phosphate synthase gene
in potato tubers perturbs the isoprenoid metabolic network: implications for
the control of the tuber life cycle. J. Exp. Bot. 57, 3007–3018.
Nisar, N., Li, L., Lu, S., Khin, N.C. and Pogson, B.J. (2015) Carotenoid
metabolism in plants. Mol Plant. 8, 68–82.
Nogueira, M., Mora, L., Enfissi, E.M., Bramley, P.M. and Fraser, P.D. (2013)
Subchromoplast sequestration of carotenoids affects regulatory mechanisms
in tomato lines expressing different carotenoid gene combinations. Plant Cell,
25, 4560–4579.
O’Connor, S.E. and Maresh, J.J. (2006) Chemistry and biology of monoterpene
indole alkaloid biosynthesis. Nat. Prod. Rep. 23, 532–547.
Ohmiya, A., Kishimoto, S., Aida, R., Yoshioka, S. and Sumitomo, K. (2006)
Carotenoid cleavage dioxygenase (CmCCD4a) contributes to white color
formation in Chrysanthemum Petals. Plant Physiol. 142, 1193–1201.
Paine, J.A., Shipton, C.A., Chaggar, S., Howells, R.M., Kennedy, M.J., Vernon,
G., Wright, S.Y., Hinchliffe, E., Adams, J.L., Silverstone, A.L. and Drake, R.
(2005) Improving the nutritional value of Golden Rice through increased provitamin A content. Nat. Biotechnol. 23, 482–487.
Rischer, H., H€akkinen, S.T., Ritala, A., Sepp€anen-Laakso, T., Miralpeix, B.,
Capell, T., Christou, P. and Oksman-Caldentey, K.M. (2013) Plant cells as
pharmaceutical factories. Curr. Pharm. Des. 19, 5640–5660.
n, M. (2010) Supply of precursors for carotenoid
Rodrıguez-Concepcio
biosynthesis in plants. Arch. Biochem. Biophys. 504, 118–122.
n, M. and Boronat, A. (2002) Elucidation of the
Rodrıguez-Concepcio
methylerythritol phosphate pathway for isoprenoid biosynthesis in bacteria
and plastids. A metabolic milestone achieved through genomics. Plant
Physiol. 130, 1079–1089.
Sambrook, J., Fritschi, E.F. and Maniatis, T. (1989) Molecular Cloning: A
Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory
Press, 1, 63–70.
Schaub, P., Al-Babili, S., Drake, R. and Beyer, P. (2005) Why is golden rice
golden (yellow) instead of red? Plant Physiol. 138, 441–450.
Schreier, P., Seftor, E., Schell, J. and Bohnert, H. (1985) The use of nuclearencoded sequences to direct the light regulated synthesis and transport of a
foreign protein into plant chloroplasts. EMBO J. 4, 25–32.
da Silva Messias, R., Galli, V., Dos Anjos, E., Silva, S.D. and Rombaldi, C.V.
(2014) Carotenoid biosynthetic and catabolic pathways: gene expression
and carotenoid content in grains of maize landraces. Nutrients, 6, 546–
563.
Sorensen, M.B., Muller, M., Skerritt, J. and Simpson, D. (1996) Hordein
promoter methylation and transcriptional activity in wild-type and mutant
barley endosperm. Mol. Gen. Genet. 250, 750–760.
Stoger, E., Williams, S., Keen, D. and Christou, P. (1999) Constitutive versus
seed specific expression in transgenic wheat: temporal and spatial control.
Transgenic Res. 8, 73–82.
Su, P.H., Yu, S.M. and Chen, C.S. (2001) Spatial and temporal expression of a
rice prolamin gene RP5 promoter in transgenic tobacco and rice. J. Plant
Physiol. 158, 247–254.
Underwood, B.A. and Arthur, P. (1996) The contribution of vitamin A to public
health. FASEB J. 10, 1040–1048.
ª 2015 Society for Experimental Biology, Association of Applied Biologists and John Wiley & Sons Ltd, Plant Biotechnology Journal, 14, 195–205
Bottlenecks in carotenoid biosynthesis in rice 205
Vallabhaneni, R., Bradbury, L.M. and Wurtzel, E.T. (2010) The carotenoid
dioxygenase gene family in maize, sorghum, and rice. Arch. Biochem.
Biophys. 504, 104–111.
Von Lintig, J. and Vogt, K. (2004) Vitamin A formation in animals: molecular
identification and functional characterization of carotene cleaving enzymes. J.
Nutr. 134, 251S–256S.
Ye, X., Al-Babili, S., Kloti, A., Zhang, J., Lucca, P., Beyer, P. and Potrykus, I.
(2000) Engineering the provitamin A (beta-carotene) biosynthetic pathway
into (carotenoids-free) rice endosperm. Science, 287, 303–305.
Zhu, C., Naqvi, S., Gomez-Galera, S., Pelacho, A.M., Capell, T. and Christou, P.
(2007) Transgenic strategies for the nutritional enhancement of plants.
Trends Plant Sci. 12, 548–555.
Zhu, C., Naqvi, S., Breitenbach, J., Sandmann, G., Christou, P. and Capell, T.
(2008) Combinatorial genetic transformation generates a library of metabolic
phenotypes for the carotenoids pathway in maize. Proc. Natl Acad. Sci. USA,
105, 18232–18237.
Supporting information
Additional Supporting information may be found in the online
version of this article:
Figure S1 Carotenoid profiles of transgenic endosperm determined by HPLC. HPLC analysis shows the profile of carotenoids
in the T3 endosperm (40 DAP) of transgenic rice lines L, D and
O (mature seeds). Line names are defined in the legend to
Figure 3. No carotenoids detected in wild type of rice
endosperm.
Table S1 Oligonucleotide sequences of forward (F) and reverse
(R) primers for mRNA blot analysis.
Table S2 Oligonucleotide sequences of forward (F) and reverse
(R) primers for quantitative real-time PCR analysis.
ª 2015 Society for Experimental Biology, Association of Applied Biologists and John Wiley & Sons Ltd, Plant Biotechnology Journal, 14, 195–205